3D tracking on transmitted-light images



Hello all,
The Trackmate XLM output file (and also the XLS’s) always include a Z-position column,
which is not used for 2D stack files. I wonder if the 3D tracking can be used
with regular transmission-light Z-stacks, or if it only works with
confocal stacks. I made an attempt and it seems to handle it, however
the spots are detected in a out-of-focus plane, notably when the
cell/blob is more white in the interior. Nevertheless, I guess this offset
shouldnt impact the results of the tracking. Am I right? How does the algorithm work for tracking on the Z-axis?
Thanks for your help!


Hi @nicolas_garcia_seyda

At the current time, the only detectors we have in TrackMate are detectors made for confocal-like images. So you need an image that is bright where the spots are.

Transmitted images are more complex. The intensity depends on the position in Z, of the contrast modality, and I do not know of a good way to retrieve the Z information in all cases from BF images (defocus rings of spots apart, but this is a specific technique).

Plus, I am surprised by your Z-stack BF images. What are they? Normally we do not acquire Z-stacks in transmitted light?

(I met an exception once however, when people were lineaging C.elegans embryo using DIC at high resolution, but again, this is a specific technique.)